INDEX

Explanations

numerical values within text

oai_token-act-pair · gpt-3.5-turbo

references to specific names or identifiers, particularly in a context of location or cultural identity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ongs

-0.93

reen

-0.80

arios

-0.78

ores

-0.77

oing

-0.76

oshenko

-0.76

uren

-0.74

cius

-0.73

ists

-0.73

ullivan

-0.73

POSITIVE LOGITS

âĢİ

0.98

 Fallen

0.82

 âĢİ

0.78

ICAN

0.77

 Barcl

0.74

lot

0.67

 tallest

0.66

 dent

0.66

STEM

0.65

":["

0.65

Activations Density 0.024%

numerical values within text

references to specific names or identifiers, particularly in a context of location or cultural identity

No Comments

No Known Activations