INDEX

Explanations

words related to providing or offering something

oai_token-act-pair · gpt-3.5-turbo

words related to affirmation and support

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

©¶æ¥µ

-0.66

 internally

-0.65

 Wonderland

-0.63

 Madness

-0.60

 subtract

-0.60

ĸļ

-0.59

GY

-0.59

²¾

-0.58

senal

-0.58

 Thinking

-0.58

POSITIVE LOGITS

irming

1.64

liction

1.55

luence

1.51

irms

1.49

ront

1.44

irmed

1.41

irmation

1.41

licted

1.37

lict

1.35

ixed

1.35

Activations Density 0.020%

words related to providing or offering something

words related to affirmation and support

No Comments

No Known Activations