INDEX

Explanations

words related to actions or events following a specific cue

oai_token-act-pair · gpt-3.5-turbo

instances of the word "then."

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

assed

-0.67

crazy

-0.65

tery

-0.64

borgh

-0.62

 Bout

-0.61

 Apart

-0.59

 Transactions

-0.59

 belief

-0.58

 fairness

-0.58

caps

-0.58

POSITIVE LOGITS

 proceeded

1.15

 proceed

0.94

Ń·

0.91

veland

0.87

 proceeds

0.84

ŃĶ

0.83

 conclud

0.81

Ļ½

0.80

©¶æ

0.77

ebin

0.77

Activations Density 0.055%

words related to actions or events following a specific cue

instances of the word "then."

No Comments

No Known Activations