INDEX

Explanations

adjectives and nouns related to comparing or contrasting features or characteristics

oai_token-act-pair · gpt-3.5-turbo

highlighted comparisons and important distinctions between concepts or ideas

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

brance

-0.89

urchase

-0.79

someone

-0.79

bsite

-0.78

onement

-0.78

ossession

-0.77

deen

-0.75

ITION

-0.75

endment

-0.72

obook

-0.72

POSITIVE LOGITS

 lessons

1.41

 factors

1.40

 trends

1.40

 reasons

1.40

 considerations

1.38

 themes

1.36

 realities

1.35

 misconceptions

1.35

 advantages

1.34

 challenges

1.33

Activations Density 0.228%

adjectives and nouns related to comparing or contrasting features or characteristics

highlighted comparisons and important distinctions between concepts or ideas

No Comments

No Known Activations