INDEX

Explanations

mentions of specific political policies and actions

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

the word "as" in various contexts and repetitions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

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_post

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

eur

-0.71

+.

-0.70

Build

-0.66

aq

-0.66

åħ

-0.64

âĸº

-0.64

+,

-0.64

aster

-0.63

itiveness

-0.63

fml

-0.63

POSITIVE LOGITS

 evidenced

1.56

 opposed

1.35

 demonstrated

1.12

 illustrated

1.10

 shown

1.04

 well

1.02

 witnessed

0.96

 exempl

0.93

 outlined

0.93

 seen

0.92

Activations Density 0.114%

mentions of specific political policies and actions

the word "as" in various contexts and repetitions

No Comments

No Known Activations