INDEX

Explanations

terms related to electronic devices

oai_token-act-pair · gpt-3.5-turbo

references to electronic devices or communication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Parables

-0.82

oons

-0.80

oglu

-0.76

eele

-0.75

cific

-0.74

hips

-0.74

oon

-0.74

aird

-0.72

xual

-0.72

aways

-0.71

POSITIVE LOGITS

 cigarettes

0.95

 warfare

0.92

 cigarette

0.90

 gadgets

0.87

 commerce

0.86

cigarette

0.82

 gadget

0.81

 devices

0.79

ronic

0.79

stasy

0.78

Activations Density 0.024%

terms related to electronic devices

references to electronic devices or communication

No Comments

No Known Activations