INDEX

Explanations

words related to the name "idd" with varying activations

oai_token-act-pair · gpt-3.5-turbo

the word "idd" indicating potential references to identity or identification

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

 metast

-0.74

pora

-0.66

utherford

-0.65

LOD

-0.62

 thin

-0.61

 tranqu

-0.61

DEM

-0.60

 retard

-0.60

 penetrating

-0.60

MER

-0.59

POSITIVE LOGITS

eenth

1.06

idd

1.00

itional

0.95

ety

0.93

olph

0.91

ick

0.90

leground

0.89

icks

0.87

oon

0.86

imensional

0.86

Activations Density 0.007%

words related to the name "idd" with varying activations

the word "idd" indicating potential references to identity or identification

No Comments

No Known Activations