INDEX

Explanations

phrases related to governmental structures and systems

oai_token-act-pair · gpt-3.5-turbo

concepts related to government and state authority

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

DEN

-0.74

Charges

-0.72

hots

-0.72

Adds

-0.71

 refresh

-0.68

OUT

-0.68

Tier

-0.68

local

-0.68

CN

-0.67

 Adren

-0.67

POSITIVE LOGITS

eers

0.89

eering

0.74

ocracy

0.74

agogue

0.73

edience

0.72

 tyranny

0.69

 bearer

0.69

alty

0.69

 morality

0.67

 apparatus

0.67

Activations Density 0.527%

phrases related to governmental structures and systems

concepts related to government and state authority

No Comments

No Known Activations