INDEX

Explanations

instances of physical obstruction or prevention

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "block" and its variations, often in the context of hindering or obstructing something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

nai

-0.71

olate

-0.70

xual

-0.69

emic

-0.69

 Volunte

-0.67

ronics

-0.67

 isEnabled

-0.66

ronic

-0.66

å°Ĩ

-0.65

rum

-0.65

POSITIVE LOGITS

 access

1.12

 entrances

1.00

 entrance

0.98

ades

0.96

 entry

0.94

aded

0.84

 passage

0.82

 exits

0.82

 doorway

0.80

 progress

0.79

Activations Density 0.091%

instances of physical obstruction or prevention

occurrences of the word "block" and its variations, often in the context of hindering or obstructing something

No Comments

No Known Activations