INDEX

Explanations

mentions of frequency or repetition

oai_token-act-pair · gpt-3.5-turbo

frequent use of the word "often" or references to repetitive occurrences

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

utenberg

-0.92

abad

-0.79

plates

-0.78

agate

-0.71

ENE

-0.70

gae

-0.70

oops

-0.70

jriwal

-0.69

eki

-0.68

 Integrity

-0.67

POSITIVE LOGITS

entimes

1.46

 overlooked

1.11

 times

0.92

 misunderstood

0.91

 referred

0.91

theless

0.89

 cited

0.89

 mistaken

0.89

 resorted

0.87

times

0.87

Activations Density 0.044%

mentions of frequency or repetition

frequent use of the word "often" or references to repetitive occurrences

No Comments

No Known Activations