INDEX

Explanations

words related to defensive positions or barriers

oai_token-act-pair · gpt-3.5-turbo

references to "guard" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

BLE

-0.80

ibal

-0.74

chy

-0.71

ml

-0.62

gom

-0.60

Cart

-0.59

article

-0.59

 PRESIDENT

-0.58

aucas

-0.58

zl

-0.56

POSITIVE LOGITS

 guard

1.03

 Guard

0.94

 guards

0.87

 Guards

0.85

guards

0.84

 guarding

0.80

ages

0.79

rail

0.79

guard

0.78

masters

0.77

Activations Density 0.014%

words related to defensive positions or barriers

references to "guard" in various contexts

No Comments

No Known Activations