INDEX

Explanations

phrases emphasizing clarification or correction

oai_token-act-pair · gpt-3.5-turbo

phrases that express caution or warnings

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

ilogy

-0.72

ourses

-0.70

ynthesis

-0.67

 stabilized

-0.66

perties

-0.65

albeit

-0.64

leground

-0.63

 stabilize

-0.63

tes

-0.62

etimes

-0.62

POSITIVE LOGITS

 yourselves

0.97

 fooled

0.87

 mistaken

0.85

 yourself

0.85

 mistake

0.82

 confused

0.79

 anymore

0.79

 confuse

0.77

 Fool

0.74

 Yourself

0.74

Activations Density 0.173%

phrases emphasizing clarification or correction

phrases that express caution or warnings

No Comments

No Known Activations