INDEX

Explanations

unique words or phrases containing 'ent'

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Ent" and its variations, indicating a focus on entities or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

å£«

-0.79

phrine

-0.71

 tremend

-0.65

 Spears

-0.65

 hearts

-0.65

CPC

-0.65

strap

-0.64

manship

-0.61

Ø©

-0.61

odcast

-0.60

POSITIVE LOGITS

ropy

1.42

rance

1.28

raction

1.20

itled

1.15

renched

1.13

reprene

1.08

rants

1.04

itle

1.03

ribute

1.02

ire

1.01

Activations Density 0.018%

unique words or phrases containing 'ent'

occurrences of the word "Ent" and its variations, indicating a focus on entities or organizations

No Comments

No Known Activations