INDEX

Explanations

references to superlatives or rankings, such as "largest" or "biggest," in various contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to large organizations, entities, or phenomena

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

utics

-0.72

adows

-0.71

 Santos

-0.69

 Jenkins

-0.68

 Oaks

-0.67

erey

-0.66

İĭ

-0.66

Laf

-0.64

 Arrows

-0.64

 Ended

-0.64

POSITIVE LOGITS

 imaginable

0.90

 ever

0.90

 mammal

0.84

 accomplishment

0.83

 EVER

0.82

Ever

0.80

 contender

0.79

 manifestation

0.77

 component

0.76

 earners

0.75

Activations Density 0.327%

references to superlatives or rankings, such as "largest" or "biggest," in various contexts

terms related to large organizations, entities, or phenomena

No Comments

No Known Activations