INDEX

Explanations

phrases indicating uncertainty or speculation

oai_token-act-pair · gpt-3.5-turbo

instances of statistical data or numeric information

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 humour

-0.83

 tornado

-0.82

 eleph

-0.82

©¶æ

-0.80

 expl

-0.80

 photograp

-0.80

 ende

-0.80

 enqu

-0.80

 thrill

-0.79

 correlated

-0.79

POSITIVE LOGITS

Critics

1.86

But

1.80

Experts

1.77

However

1.76

SPONSORED

1.75

Officials

1.74

Yet

1.73

Story

1.72

Instead

1.71

Nevertheless

1.70

Activations Density 0.369%

phrases indicating uncertainty or speculation

instances of statistical data or numeric information

No Comments

No Known Activations