INDEX

Explanations

references to monetary values expressed in British pounds

oai_token-act-pair · gpt-3.5-turbo

references to monetary values

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

defense

-0.86

avored

-0.85

 behaviors

-0.83

 behavior

-0.78

 utilized

-0.75

 organized

-0.74

 honoring

-0.74

 counselors

-0.74

 organizations

-0.73

 demeanor

-0.72

POSITIVE LOGITS

 Â£

4.01

Â£

3.15

 âĤ¬

2.85

 (Â£

2.40

Rs

2.19

 Â¥

1.97

âĤ¬

1.87

1.80

NHS

1.40

UK

1.39

Activations Density 0.011%

references to monetary values expressed in British pounds

references to monetary values

No Comments

No Known Activations