INDEX

Explanations

license plate numbers and identifiers

oai_token-act-pair · gpt-3.5-turbo

references to grant numbers and license plate formats

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 wiser

-0.72

 exha

-0.72

 disproportion

-0.72

 clos

-0.68

 behavi

-0.66

 healthier

-0.66

ividual

-0.66

 bowls

-0.65

emouth

-0.65

 destro

-0.64

POSITIVE LOGITS

1.55

1.53

1.45

1.44

1.43

1.41

Activations Density 0.152%

license plate numbers and identifiers

references to grant numbers and license plate formats

No Comments

No Known Activations