INDEX

Explanations

words related to barriers or obstacles

oai_token-act-pair · gpt-3.5-turbo

the word "penetrate" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

FY

-0.82

 coffee

-0.78

 attendant

-0.74

mer

-0.71

 summary

-0.67

 espresso

-0.65

 recap

-0.64

MSM

-0.63

 sampling

-0.62

 integration

-0.61

POSITIVE LOGITS

etr

4.40

 Petr

1.18

yth

1.08

etrical

1.08

emet

0.99

egu

0.94

epad

0.92

umbered

0.90

eline

0.90

rypt

0.90

Activations Density 0.020%

words related to barriers or obstacles

the word "penetrate" and its variations

No Comments

No Known Activations