INDEX

Explanations

phrases indicating a hypothetical situation or condition

oai_token-act-pair · gpt-3.5-turbo

conditional phrases indicating hypothetical scenarios

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Els

-0.66

ãĥ³ãĤ¸

-0.62

Alz

-0.62

 Eisen

-0.61

annels

-0.61

Ts

-0.61

ilib

-0.60

EStreamFrame

-0.58

Flor

-0.57

 Wick

-0.57

POSITIVE LOGITS

 someday

0.83

 succeed

0.80

 somehow

0.79

 succeeds

0.79

ado

0.72

acles

0.71

rha

0.71

 mirac

0.70

 properly

0.68

chance

0.68

Activations Density 0.897%

phrases indicating a hypothetical situation or condition

conditional phrases indicating hypothetical scenarios

No Comments

No Known Activations