INDEX

Explanations

phrases related to political statements and actions

oai_token-act-pair · gpt-3.5-turbo

phrases highlighting actions and comments made by specific individuals or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

chwitz

-0.84

culosis

-0.79

hazard

-0.75

emn

-0.75

eno

-0.75

tle

-0.72

MpServer

-0.72

mol

-0.71

 resil

-0.71

iamond

-0.70

POSITIVE LOGITS

 bureaucrats

1.13

 politicians

1.07

 authorities

0.91

 prosecutors

0.90

 judges

0.88

 individuals

0.87

 governors

0.87

 legislators

0.86

 regulators

0.85

 businessmen

0.85

Activations Density 0.310%

phrases related to political statements and actions

phrases highlighting actions and comments made by specific individuals or groups

No Comments

No Known Activations