INDEX

Explanations

phrases related to controversy or opposition

oai_token-act-pair · gpt-3.5-turbo

markers of negative judgment or criticism

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

enta

-0.73

 Bunny

-0.67

ctors

-0.65

 Somerset

-0.65

 unmarked

-0.63

 bunny

-0.62

creen

-0.62

 sled

-0.62

dig

-0.62

 Manhattan

-0.61

POSITIVE LOGITS

âĹ¼

0.97

âĢł

0.90

0.86

Â§

0.84

âģ

0.81

ISIS

0.81

Â¯

0.80

0.78

¬

0.77

â

0.75

Activations Density 0.358%

phrases related to controversy or opposition

markers of negative judgment or criticism

No Comments

No Known Activations