INDEX

Explanations

political keywords related to specific individuals or groups

oai_token-act-pair · gpt-3.5-turbo

references to political figures and entities associated with ideological divides

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

oscope

-0.75

 Adren

-0.71

phies

-0.71

pak

-0.70

redits

-0.66

aps

-0.65

effective

-0.65

osc

-0.62

pipe

-0.62

aft

-0.61

POSITIVE LOGITS

istani

0.81

anmar

0.77

 hybrids

0.72

 unity

0.72

heid

0.71

 sentiment

0.71

 relations

0.70

iosity

0.70

 sentiments

0.70

 hierarch

0.69

Activations Density 0.102%

political keywords related to specific individuals or groups

references to political figures and entities associated with ideological divides

No Comments

No Known Activations