INDEX

Explanations

phrases emphasizing the absence or negation of something

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing the lack of evidence or the universal applicability of a statement

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

Module

-0.61

pat

-0.61

NES

-0.60

elfth

-0.59

ÅĤ

-0.59

emn

-0.59

rig

-0.58

itely

-0.57

erb

-0.54

kers

-0.53

POSITIVE LOGITS

 except

0.76

 else

0.74

 alike

0.72

0.70

agher

0.68

--

0.68

 imaginable

0.67

soever

0.67

ãĢĤ

0.66

0.65

Activations Density 0.153%

phrases emphasizing the absence or negation of something

phrases emphasizing the lack of evidence or the universal applicability of a statement

No Comments

No Known Activations