INDEX

Explanations

instances where something is considered "too much" or "excessive"

oai_token-act-pair · gpt-3.5-turbo

phrases expressing the concept of excess or being "too much" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

riots

-0.72

Script

-0.68

eers

-0.66

creator

-0.66

Russ

-0.65

ords

-0.65

 Siber

-0.65

inators

-0.65

-0.64

eer

-0.64

POSITIVE LOGITS

 risky

0.74

 busy

0.74

ooo

0.72

 distracting

0.71

 costly

0.69

oths

0.69

 obvious

0.68

 much

0.68

 difficult

0.67

 tempting

0.67

Activations Density 0.036%

instances where something is considered "too much" or "excessive"

phrases expressing the concept of excess or being "too much" in various contexts

No Comments

No Known Activations