INDEX

Explanations

expressions related to actions or events

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions, particularly those associated with significant events or reactions

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

Ħ¢

-0.58

 anymore

-0.57

stones

-0.50

parent

-0.50

?".

-0.49

bits

-0.49

heed

-0.49

".

-0.49

ny

-0.48

%.

-0.48

POSITIVE LOGITS

 last

0.61

 yesterday

0.58

 earlier

0.57

 aback

0.55

çīĪ

0.54

 enthusi

0.54

Roz

0.54

velt

0.53

 Uriel

0.51

 extensively

0.50

Activations Density 1.810%

expressions related to actions or events

phrases indicating actions, particularly those associated with significant events or reactions

No Comments

No Known Activations