INDEX

Explanations

ways to take action or respond to situations

oai_token-act-pair · gpt-3.5-turbo

keyword phrases related to reactions or responses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

alist

-0.46

"))

-0.45

anism

-0.43

ye

-0.43

frac

-0.42

division

-0.41

ATIONAL

-0.41

)].

-0.41

)",

-0.40

asking

-0.39

POSITIVE LOGITS

 however

0.86

 meanwhile

0.61

 moreover

0.58

 therefore

0.58

 though

0.57

 ital

0.50

 furthermore

0.50

Nek

0.44

 accordingly

0.44

tho

0.44

Activations Density 2.359%

ways to take action or respond to situations

keyword phrases related to reactions or responses

No Comments

No Known Activations