INDEX

Explanations

phrases related to pairs or sets of objects or entities

oai_token-act-pair · gpt-3.5-turbo

references to competitive entities or groups within various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

taboola

-0.76

humane

-0.66

ople

-0.65

tun

-0.62

etc

-0.60

phia

-0.60

pron

-0.60

nance

-0.59

pmwiki

-0.59

azard

-0.58

POSITIVE LOGITS

 apiece

1.10

 totaling

1.01

 respectively

1.01

 simultaneously

0.97

 concurrently

0.90

 consecut

0.88

 collided

0.87

 vying

0.85

 jointly

0.83

 together

0.80

Activations Density 0.450%

phrases related to pairs or sets of objects or entities

references to competitive entities or groups within various contexts

No Comments

No Known Activations