INDEX

Explanations

instances of imperative statements related to potential actions or choices

oai_token-act-pair · gpt-3.5-turbo

phrases related to purchasing or shopping decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

itudinal

-0.73

 Actions

-0.73

 moder

-0.73

leaders

-0.73

voice

-0.72

 Reconstruction

-0.70

 largeDownload

-0.70

 Leaders

-0.70

 Voting

-0.69

 Acting

-0.69

POSITIVE LOGITS

 purchase

1.84

buy

1.75

 purchases

1.48

 buying

1.47

 buys

1.45

 purchasing

1.45

 purchased

1.36

 procure

1.35

 bought

1.35

buy

1.33

Activations Density 0.472%

instances of imperative statements related to potential actions or choices

phrases related to purchasing or shopping decisions

No Comments

No Known Activations