INDEX

Explanations

proper nouns related to entertainment, particularly movies or TV shows

oai_token-act-pair · gpt-3.5-turbo

character names and related elements from a narrative context

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

poke

-0.62

 Kislyak

-0.62

Offline

-0.62

thirds

-0.60

elo

-0.59

wcsstore

-0.59

 tbsp

-0.59

dos

-0.58

 dice

-0.56

ItemTracker

-0.56

POSITIVE LOGITS

vation

0.73

irie

0.63

oire

0.63

ament

0.62

heid

0.61

hyde

0.61

gerald

0.60

.?

0.60

.:

0.59

aments

0.59

Activations Density 0.074%

proper nouns related to entertainment, particularly movies or TV shows

character names and related elements from a narrative context

No Comments

No Known Activations