INDEX

Explanations

occurrences of the word "too" followed by a number as an indication of excessiveness in various contexts

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "too" in various contexts indicating excess or overuse

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

hyde

-0.75

idia

-0.74

quist

-0.68

inarily

-0.68

psons

-0.64

icipated

-0.64

eer

-0.64

Script

-0.63

issance

-0.62

eval

-0.61

POSITIVE LOGITS

ħĭ

0.73

 risky

0.71

 stren

0.70

oths

0.70

 distracting

0.69

lax

0.68

 tempting

0.66

 much

0.66

 abst

0.65

 trusting

0.65

Activations Density 0.037%

occurrences of the word "too" followed by a number as an indication of excessiveness in various contexts

occurrences of the word "too" in various contexts indicating excess or overuse

No Comments

No Known Activations