INDEX

Explanations

verbs ending in '-ided' which might indicate a state of self-disclosure or sharing personal experiences

oai_token-act-pair · gpt-3.5-turbo

forms of the word "decide" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

adra

-0.69

thur

-0.69

thia

-0.69

iona

-0.66

ctica

-0.65

SEA

-0.63

apo

-0.63

isma

-0.61

oen

-0.61

chrom

-0.61

POSITIVE LOGITS

escription

0.87

ided

0.83

own

0.79

aily

0.78

ãĥ¼ãĤ¯

0.75

boarding

0.74

maid

0.73

ide

0.72

 Franch

0.71

ividual

0.71

Activations Density 0.010%

verbs ending in '-ided' which might indicate a state of self-disclosure or sharing personal experiences

forms of the word "decide" in various contexts

No Comments

No Known Activations